Walking speed and stride length predicts 36 months dependency, mortality, and institutionalization in Chinese aged 70 and older

BACKGROUND: Increasing emphasis is being placed on physical performance measures as an outcome predictor. It is uncertain whether one or two simple measurements will have predictive value compared with a battery of tests. OBJECTIVES: To assess whether simple performance measures such as walking speed and stride length will predict dependency, mortality, and institutionalization. DESIGN: A 3-year longitudinal study of a random sample of subjects. SETTING: Older people living in the community in Hong Kong, Special Administrative Region, China. SUBJECTS: A total of 2032 Chinese subjects aged 70 years and older were recruited territory-wide by proportional random sampling and followed for 3 years. MEASUREMENTS: Functional status was measured using the Barthel Index at baseline and follow-up. The time taken to walk a distance of 16 feet and the number of steps taken were measured at baseline. Stride length is estimated by dividing 16 by the average number of steps needed to complete the walk. Outcomes regarding dependency, mortality, and institutionalization at 36 months were recorded. RESULTS: After excluding subjects lost to follow-up and those who had died, data were available for 559 men and 612 women. Univariate analysis showed that reduced walking speed and stride length were associated with increased risk of dependency, mortality, and institutionalization. In multivariate analysis for dependency and mortality, stride length, walking speed, age, and sex were included in the best prediction model (ROC = 0.798 and 0.707, respectively), whereas only stride length was included in the prediction for institutionalization (ROC = 0.764). CONCLUSIONS: In terms of prevention or modifying outcomes, these two simple performance measures may be used as indicators for checking for occult disease and for interventional measures such as exercise prescription.